Skip to contents

Evaluate Local Continuity Meta-Criterion (LCMC)

lcmc.Rd

Evaluate Local Continuity Meta-Criterion (LCMC)

Usage

lcmc(x, y, ks, fit = NULL)

Arguments

x

Raw data/embeddings in original space.

y

Raw data/embeddings in new (dimension-reduced) space.

ks

Vector of neighborhood sizes to evaluate LCMC.

fit

(Optional) Fitted locoMDS object corresponding to the embeddings in y. Only used for warning messages about uniqueness of points.

Value

A data frame with columns:

  • k: Neighborhood size.

  • lcmc: Local Continuity Meta-Criterion value for the given neighborhood size.

  • adjusted_lcmc: Adjusted LCMC value, accounting for the neighborhood size.

Examples

# generate example data
n <- 100
p1 <- 4
p2 <- 2
x <- matrix(rnorm(n * p1), nrow = n, ncol = p1)
y <- matrix(rnorm(n * p2), nrow = n, ncol = p2)

# evaluate LCMC for neighborhood sizes 5, 10, and 20
lcmc_out <- lcmc(x, y, ks = c(5, 10, 20))